Frequency Response

Ideal frequency response - solid green line (1).

Common response errors are shown by dashed lines.

2 - under damped bass peak gives lively, 'obvious' bass.

3 - over damped bass gives 'tight' bass and suits near wall placement.

4 - a small up / down blip suggests a strong internal reflection and resultant colouration.

5 - a dip around 3kHz is due to driver mismatch in the crossover and softens the sound.

6 - raised output from the tweeter, above 3kHz, results in a bright sound with strong detail.

7 - treble peak at 15kHz produces sharp treble.

8 - falling output from the tweeter, above 3kHz, results in warm sound.

WHAT IT TELLS US

Frequency response is a guide to tonal balance, or whether cymbals will burn your ears out whilst kick drum lacks kick. Ideally, the green trace of our frequency response graph should be horizontal within a few dB (decibels)  from 100Hz to 20kHz, as in the solid green line above. It should also be flat and smooth, as this indicates a lack of resonances. Dome tweeters in particular, which typically work from 3kHz up to 20kHz, can look ragged and uneven in their response (see 7 above) and sound coarse and coloured as a result, whilst better quality ribbon tweeters (and others such as magneto planars and ring domes)  measure almost ruler flat and sound commensurately smoother, less coloured and more natural. Frequency response tells quite a big story and, with loudspeakers, the bumpier it is the more colour there will be.

To make loudspeakers sound detailed and clear in showroom demonstration against rivals, a majority have raised output above 3kHz (as in 6 above). Taken too far, this results in a fatiguingly sharp sound in the home, especially when a poor metal dome tweeter dome is used. Our graphs warn of this common ‘enhancement’, that’s really a blight.

At low frequencies it is common to acoustically under damp bass units to give them a little life and presence at low volumes and make bass content stand out (see 2 above). Again, how acceptable this is is a matter of degree. The effect can be seen as a small peak of a few dB, usually around 80Hz, in the green trace of our graphs and about +2dB is all that is acceptable. The response we publish slightly under states this effect; our unpublished pink noise analysis makes this effect clearer. Too much bass lift does mean bass will sound boomy; there is strong correlation here.

Some loudspeakers are over-damped and have progressively falling bass (see 3 above). They work well with powerful amplifiers, giving solid punchy lows, free of waffle, but only at medium to high volume. They are good for Rock played loud. This type of loudspeaker works well close to or against a rear wall. Small bookshelf and wall mounting loudspeakers and even floor standers like the KEF iQ5, designed for near wall use, commonly produce less bass to compensate for rear wall presence.

There are arguments in favour of a non-flat forward response, such as maintaining constant acoustic power and compensating for baffle effects. However, subjectively they upset the forward, direct sound to a listener in quite a strong manner and are rarely pursued.

Epos Encore 50 frequency response - green trace. Port - red trace.

A majority of modern loudspeakers have ports (bass reflex). The port both damps resonance in the bass unit(s) and extends output downward. We show port output in the red trace (see above). A narrow tuned port will give extra bounce to the bass, but not “tightness”. Broad tuned ports, a recent development, apply more damping to the bass unit and this usefully reflects back into the electrical load, making the  loudspeaker more amplifier friendly. Broad tuned ports do result in more controlled sounding bass.

Also, to keep the loudspeaker sounding fairly spry in its bass behaviour, companies like B&W and KEF now tune the port high, to 40Hz, the lower end of the bass range. This eliminates subsonics, or rumbly effects, but it does make for very tidy sounding bass lines. So the red port trace tells quite a big story.

HOW WE MEASURE IT

We make an initial measurement using third octave analysis of pink noise, moving the microphone to find a reasonably representative microphone position for response measurement, and also assess off axis behaviour and poor phase matching between drive units.

Epos Encore 50 third-octave analysis of pink noise.

Our published response curve is made using  gated sine wave bursts, with delay compensation. Note that this system as used by Clio weights down low frequency data below 180Hz in the FFT measurement window so there is progressive loss of detail in the plots and they are lower in amplitude than the steady state data collected using pink noise; compare the two Encore 50 plots above to see this. However, they are close to more laboriously collected, spliced-in near field data, which gives a true ‘anechoic’ result, differences being small. Comparison of results and discussion with a large number of loudspeaker manufacturers makes us confident that our published frequency response data is accurate in itself and representative of a product’s performance.

The red port trace is not positioned to match the green trace in terms of Sound Pressure Level, or Acoustic Power. However, we do measure SPL at the port entrance relative to that at the drive unit (nearfield) at 80Hz to provide guidance as to relative levels, and quote the figure in our Measured Performance summary. In practice ports commonly measure around +6dB above drive unit output at 80Hz. Large ports exceed this and small ones provide less output.